Received: 9 August 2018 

|
  Revised: 14 January 2019 
|  Accepted: 10 March 2019

DOI: 10.1111/eci.13101

ORIGINAL ARTICLE

eGFR and coronary artery calcification in chronic kidney disease

Young Youl Hyun1   | Hyang Kim1  | Kook‐Hwan Oh2  | Curie Ahn2  | Sue K. Park3  |

Dong Wan Chae4  | Yun Kyu Oh5  | Kyu Hun Choi6  | Seung Hyeok Han6  |
Yeong Hoon Kim7  | Kyu‐Beck Lee1 KNOW-CKD Study Investigators

1
Department of Internal
Medicine, Sungkyunkwan University
Abstract
School of Medicine, Kangbuk Samsung Background: The independent association between eGFR and coronary artery calci-
Hospital, Seoul, Korea fication (CAC) is complex and not clear. The aim of this study was to investigate the
2
Department of Internal Medicine, Seoul
relationship between eGFR calculated from different equations and CAC in predialy-
National University College of Medicine,
Seoul, Korea sis CKD patients in Korea.
3
Department of Preventive Medicine, Seoul Methods: In this cross‐sectional study, we analysed 1533 patients from the KNOW‐
National University College of Medicine, CKD cohort. eGFR was calculated by a four‐variable MDRD equation (eGFRMDRD),
Seoul, Korea
4
CKD‐EPI creatinine equations (eGFRCr), CKD‐EPI cystatin C equation (eGFRCys)
Department of Internal Medicine, Seoul
National University Bundang Hospital, and CKD‐EPI creatinine‐cystatin equation (eGFRCrCys). Participants were divided
Seongnam, Gyeonggi‐do, Korea into eGFR categories (<30, 30‐59, 60‐89, ≥90 mL/min/1.73 m2). CACS (coronary
5
Department of Internal Medicine, Seoul artery calcium score) was measured using cardiac computed tomography. CAC was
National University Boramae Medical
defined as CACS >100.
Center, Seoul, Korea
6
Department of Internal Medicine, Yonsei
Results: Coronary artery calcification was found in 334 (21.8%) patients and was
University College of Medicine, Seoul, more prevalent in the lower eGFR groups (P < 0.001). In multivariate Tobit regres-
Korea sion, CACS increased gradually as eGFRCrCys decreased (P for trend = 0.034). In
7
Department of Internal Medicine, Busan
multivariate logistic regression, there were gradual associations between lower eGFR
Paik Hospital, Inje University, Busan,
Korea and CAC when an eGFRCys or eGFRCrCys was used. The adjusted OR for CAC in the
group with eGFR <30 mL/min/1.73 m2 compared to the group with eGFR ≥90 mL/
Correspondence
min/1.73 m2 was 2.64 (95% CI, 1.09‐3.60) when eGFRCrCys was used. Of the four
Kyu‐Beck Lee, Department of Internal
Medicine, Division of Nephrology, eGFR formulas, only adding eGFRCrCys significantly improved CAC prediction
Kangbuk Samsung Hospital, models without eGFR (P = 0.046).
Sungkyunkwan University School of
Medicine, Seoul, Korea.
Conclusions: There was a gradual and independent association between low eGFR
Email: kyubeck.lee@samsung.com and CAC in a predialysis CKD cohort in Korea. eGFRCrCys predicted CAC better
Funding information
than other equations in this population.
This work was supported by the Research
Program funded by the Korea Centers KEYWORDS
for Disease Control and Prevention chronic kidney disease, coronary artery calcification, eGFR, renal function
(2011E3300300, 2012E3301100,
2013E3301600, 2013E3301601,
2013E3301602, 2016E3300200,
2016E3300201).

© 2019 Stichting European Society for Clinical     1 of 8

Eur J Clin Invest. 2019;49:e13101. wileyonlinelibrary.com/journal/eci |
https://doi.org/10.1111/eci.13101 Investigation Journal Foundation
 |
2 of 8       HYUN et al.
1  |   IN TRO D U C T ION
Coronary artery calcification (CAC), a marker of atheroscle-
rosis, is a well‐known predictor of frequent cardiovascular
events and high mortality in the general population1,2 and in
patients with chronic kidney disease (CKD).3-6 Vascular cal-
cification is an important clinical manifestation of CKD and
is considered to be a mediator of high cardiovascular mortal-
ity in CKD patients.
The prevalence of CAC is much higher in CKD patients
than in the general population whether they are on dialysis
or not.7-10 Given this fact, it is easy to assume that CKD is a
risk for CAC and that lower eGFR is associated with more
severe CAC in predialysis CKD patients. However, it is more
complicated than we think. In previous studies, important
risk factors such as phosphorus were not included in the anal-
ysis, or the association between low eGFR and CAC was dis-
appeared when the analysis was adjusted for risk factors for
CAC. Therefore, it is unclear whether the cause of CAC in
low eGFR is comorbid diseases or the decreased renal func-
tion itself. In addition, most of the existing studies on eGFR
and CAC used the Modification of Diet in Renal Disease
(MDRD) Study equation to estimate GFR.11 However, recent
studies have shown that eGFR calculated from the Chronic
Kidney Disease Epidemiology Collaboration (CKD‐EPI)
equation is more relevant to real GFR and predicts the out-
come more accurately.12,13 However, it is unclear which of
the various eGFR formulas better predicts CAC in patients
with CKD.
We hypothesized that there would be an independent
associations between eGFR and CAC in CKD patients and
that CKD‐EPI equation might be better in predicting CAC.
The aim of this study was to investigate the relationship
between eGFR and CAC in a predialysis CKD cohort of
Korean adults. This analysis will include traditional and
nontraditional cardiovascular risk factors and use different
eGFR formulas.
2  |  M AT E R IA L A N D ME T H O DS
2.1  |  Study design and population
This is a cross‐sectional study designed to examine the as-
sociation between eGFR and CAC in CKD patients. The
study subjects were participants in the prospective Korean
Cohort Study for Outcome in Patients With Chronic Kidney
Disease (KNOW‐CKD). A detailed protocol of this study
has been published previously.14 The study protocol was
approved by the Institutional Review Board at each par-
ticipating clinical centre in 2011. In brief, KNOW‐CKD is
a prospective cohort study that enrols subjects with predi-
alysis CKD stages 1‐5 and who are between the ages of 20
and 75 years. Nine nephrology centres in major university
hospitals throughout Korea enrolled 2238 adults with
chronic kidney disease over a 5‐year period, from 2011 to
2015. The participating individuals will be monitored for
approximately 10 years until death or until end‐stage renal
disease occurs.
Among the cohort, 2113 patients who had undergone
coronary multi‐detector computed tomography (MDCT)
were potential study participants. We first excluded 496
patients for whom we were missing data for the variables
of interest. Then, we excluded 84 patients with previous
coronary artery disease. The final analysis included 1533
patients (Figure 1).
2.2  |  Clinical and laboratory measurements
Data on sociodemographic information, medical his-
tory, medication use and health‐related behaviour were
collected by means of a self‐administered questionnaire,
with the assistance of trained staff. Anthropometric data
and resting blood pressure (BP) were measured by trained
nurses. Blood samples were taken after fasting for at least
10 hours. Random urine samples from midstream collec-
tion were used to measure the urine protein‐to‐creatinine
ratio (PCR). Serum creatinine, cystatin C and 25‐(OH)‐vit
D were measured at the central laboratory. Other biochemi-
cal analyses were done at the local laboratory of each par-
ticipating centre. Serum creatinine levels were measured by
the isotope dilution mass spectroscopy (IDMS)‐traceable
method. Hypertension was defined as a systolic blood pres-
sure >140 mm Hg, diastolic blood pressure >90 mm Hg or
a history of hypertension. Diabetes mellitus was defined
as a fasting serum glucose >126 mg/dL or a history of
diabetes.
KNOW-CKD baseline cohort (n = 2238)
Exclusion
• Patients without coronary CT (n =125)
KNOW-CKD cohort with CACS (n = 2113)
Exclusion
• Missing data (n = 496)
Participants without missing data (n = 1617)
Exclusion
• Previous coronary artery disease (n = 84)
Final study subjects (n = 1533)
F I G U R E 1   Algorithm for selecting study subjects from the
KNOW‐CKD cohort. CACS, coronary artery calcium score; CT,
computed tomography
HYUN et al.   
|
    3 of 8

2.3  |  Estimation of GFR 3  |  RESULTS

eGFR was calculated by the four‐variable Modification of
Among 1533 study subjects, a total of 334 (21.8%) partici-
Diet in Renal Disease (MDRD) equation (eGFRMDRD)11 and
pants had CAC. The characteristics of participants in terms
the CKD‐EPI creatinine equations (eGFRCr),12 CKD‐EPI
of the eGFRCrCys categories are compared in Table 1. The
cystatin C equation (eGFRCys) and CKD‐EPI creatinine‐cys-
lower the eGFR, the higher the CACS. The median values
tatin equation (eGFRCrCys).15
of CACS from the ≥90 mL/min/1.73 m2 category to the
<30 mL/min/1.73 m2 category were 0.0, 0.0, 4.3 and 9.5,
2.4  |  Measurement of CAC respectively (P < 0.001). Similarly, CAC was more preva-
lent in the lower eGFR groups (4.1%, 16.0%, 25.9% and
Electrocardiography‐gated coronary MDCT scans were car-
31.2% for the ≥90 mL/min/1.73 m2 category compared to
ried out following the standard protocol of each centre. The
the <30 mL/min/1.73 m2 category of eGFR, P < 0.001). As
quantitative CACS was calculated according to the method
expected, participants with lower eGFR had worse cardio-
described by Agatston et al16 The presence of CAC was de-
vascular profiles. For example, they had higher age, WHR,
fined as a CACS >100 in the present study. A CACS >100
systolic and diastolic BP, serum glucose, LDL cholesterol,
differentiates between mild and moderate‐to‐severe coronary
hsCRP, phosphorus, and iPTH. Diabetes and hypertension
artery disease.1,17,18 The associations using CAC with differ-
were more common in participants with lower eGFR. The
ent cut‐offs (CACS > 10, CACS > 50, and CACS > 200)
distribution of eGFR values from four eGFR equations is
were also evaluated in sensitivity analysis.
shown in the Figure S1.
Multivariable‐adjusted CACS ratios according to
2.5  |  Statistical analysis eGFR categories were calculated from Tobit regression
(Table 2). Although adjustment for more risk factors at-
Participants were divided into eGFR categories (<30, 30‐59,
tenuated the association, the group with eGFR <30 mL/
60‐89, ≥90 mL/min/1.73 m2). Continuous variables were
min/1.73 m2 showed significantly higher CACS than the
expressed as the mean ± SD or median (interquartile range).
group with eGFR ≥90 mL/min/1.73 m2 regardless of
Continuous variables were compared between two groups
the eGFR formula. Especially, when the eGFRCrCys for-
using a t test or the Mann‐Whitney U test and between four
mula was used, there was a gradual correlation showing
groups using analysis of variance or the Kruskal‐Wallis
that the lower the eGFR, the higher the CACS (P for
test. Categorical variables were expressed as percentages
trend = 0.034).
and compared between groups using the chi‐square test.
Table 3 summarizes the results of multivariate logis-
To evaluate the association between eGFR categories and
tic regression analysis and shows the association between
CACS, we used a Tobit regression model for natural log
eGFR and CAC. There were gradual associations between
(CACS + 1). Tobit models were used to estimate ratios
lower eGFR and CAC when the cystatin C‐based eGFR
and 95% confidence intervals (CIs) of CACS + 1 compar-
equation (eGFRCys or eGFRCrCys) was used to estimate
ing lower eGFR categories with ≥90 mL/min/1.73 m2 eGFR
renal function, but not when eGFRMDRD or eGFRCr was
category. We further used logistic regression analysis to cal-
used. The adjusted OR for CAC in the group with eGFR
culate the odds ratio (OR) and CI for the presence of CAC
<30 mL/min/1.73 m2 compared to the group with eGFR
associated with lower eGFR categories compared to the
≥90 mL/min/1.73 m2 was 2.64 (95% CI, 1.09‐3.60) when
≥90 mL/min/1.73 m2 eGFR category. Multivariate models
eGFRCrCys was used. When each eGFR using four formulas
were adjusted for age, sex, waist‐hip ratio (WHR), systolic
was added to CAC prediction models without eGFR, only
BP, diabetes, LDL (low‐density lipoprotein) cholesterol,
eGFRCrCys improved the model with statistical significance
HDL (high‐density lipoprotein) cholesterol, high‐sensitivity
(eGFRMDRD, P = 0.246; eGFRCr, P = 0.267; eGFRCys,
C‐reactive protein (hsCRP), urine PCR, statin use, smoking
P = 0.061; eGFRCrCys, P = 0.046).
status, serum calcium, serum phosphorus, 25‐OH‐vit D and
Our study defined CAC as a CACS >100, but the asso-
intact parathyroid hormone (iPTH). To test the improvement
ciation between eGFR and CAC with different cut‐offs was
of the CAC prediction model by adding eGFR, a likelihood
also evaluated in sensitivity analysis (Table 4). Compared to
ratio test was performed. Statistical analyses were performed
the group with eGFR ≥90 mL/min/1.73 m2, the group with
using Stata version 15.0 (StataCorp LP, College Station, TX,
eGFR <30 mL/min/1.73 m2 had an OR for CAC of 2.53
USA). Reporting of the study conforms to STROBE state-
(95% CI, 1.02‐6.26; P = 0.045) when CAC was defined as
ment along with references to STROBE statement and the
a CACS >200. The results were similar when CAC was de-
broader EQUATOR guidelines.19
fined as CACS >0 or CACS >50.
 |
4 of 8       HYUN et al.

T A B L E 1   Baseline characteristics in relation to eGFRCrCys

eGFRCrCys (mL/min/1.73 m2)

Characteristics Total ≥90 60‐89 30‐59 <30 P for trend

N 1533 271 282 525 455  
CACS 0.0 (0.0‐71.4) 0.0 (0.0‐0.0) 0.0 (0.0‐30.0) 4.3 (0.0‐109.0) 9.5 (0.0‐167.5) <0.001
CACS > 100 334 (21.8) 11 (4.1) 45 (16.0) 136 (25.9) 142 (31.2) 0.099
Age (y) 53.2 ± 12.2 43.1 ± 11.2 51.2 ± 11.2 55.7 ± 11.2 57.4 ± 10.9 <0.001
Male sex 917 (59.8) 131 (48.3) 188 (66.7) 335 (63.8) 263 (57.8) 0.099
2
BMI (kg/m ) 24.6 ± 3.5 24.3 ± 3.7 24.6 ± 3.5 24.7 ± 3.3 24.4 ± 3.6 0.721
WHR 0.896 ± 0.066 0.874 ± 0.066 0.891 ± 0.064 0.904 ± 0.064 0.905 ± 0.066 <0.001
Systolic BP 128 ± 16 126 ± 14 126 ± 15 126 ± 15 131 ± 18 <0.001
(mm Hg)
Diastolic BP 77 ± 11 79 ± 11 78 ± 11 76 ± 10 77 ± 12 0.008
(mm Hg)
Glucose (mg/dL) 109 ± 38 101 ± 21 107 ± 33 112 ± 39 114 ± 46 <0.001
LDL cholesterol 98 ± 31 107 ± 31 102 ± 31 95 ± 30 92 ± 32 <0.001
(mg/dL)
HDL cholesterol 50 ± 15 57 ± 16 53 ± 15 49 ± 15 45 ± 14 <0.001
(mg/dL)
hsCRP (mg/dL) 0.6 (0.2‐1.7) 0.4 (0.1‐1.2) 0.6 (0.3‐1.7) 0.6 (0.2‐1.7) 0.8 (0.3‐2.1) <0.001
eGFR (mL/ 54 ± 33 110.1 ± 12.5 73.7 ± 8.8 44 ± 8.8 19.9 ± 6 <0.001
min/1.73 m2)
Calcium (mg/dL) 9.1 ± 0.5 9.3 ± 0.4 9.3 ± 0.4 9.2 ± 0.5 8.9 ± 0.6 <0.001
Phosphorus (mg/ 3.7 ± 0.7 3.5 ± 0.5 3.5 ± 0.6 3.6 ± 0.5 4.1 ± 0.8 <0.001
dL)
25‐(OH)‐vit D (ng/ 17.5 ± 7.6 16.6 ± 6.6 18.5 ± 7.3 18 ± 7.9 17 ± 7.9 0.974
mL)
Intact PTH (pg/mL) 51 (33‐84) 35 (25‐47) 36 (27‐51) 51 (36‐74) 96 (63‐154) <0.001
Urine PCR (g/g) 0.5 (0.2‐1.6) 0.2 (0.1‐0.6) 0.3 (0.1‐0.9) 0.5 (0.2‐1.5) 1.2 (0.5‐3) <0.001
Hypertension 1482 (96.7) 238 (87.8) 273 (96.8) 519 (98.9) 452 (99.3) <0.001
Diabetes 521 (34.0) 42 (15.5) 67 (23.8) 197 (37.5) 215 (47.3) <0.001
Statin use 768 (50.1) 95 (35.1) 134 (47.5) 290 (55.2) 249 (54.7) <0.001
Current smoker 258 (16.8) 35 (12.9) 54 (19.2) 91 (17.3) 78 (17.1) 0.301
BMI, body mass index; BP, blood pressure; CAC, coronary artery calcification; CACS, coronary calcium score; eGFR, estimated glomerular filtration rate; HDL, high‐
density lipoprotein; hsCRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; PCR, protein‐to‐creatinine ratio; PTH, parathyroid hormone; PWV, pulse
wave velocity; TC, total cholesterol; WHR, waist‐hip ratio.
Values for categorical variables are reported as a number (percentage); and for continuous variables, as mean ± SD or median [interquartile range]. P values are for
analysis of variance or Kruskal‐Wallis tests for continuous variables and chi‐square tests for categorical variables.

and iPTH. In addition, this association was robust against

4  |   D IS C U SS ION different CACS cut‐offs for CAC. Adding eGFRCrCys signifi-
cantly improved the CAC prediction model without eGFR.
Among the 1533 patients in the predialysis CKD cohort
Given previous studies on the association of CKD
of Korean adults who had available baseline CACS, lower
or renal function with CAC, we tend to accept that
eGFR was gradually associated with higher CACS and
low renal function is undoubtedly a risk for CAC, but
higher prevalence of CAC. This association was independent
the real relationship is not so clear. Kramer et al 10 an-
of traditional and nontraditional cardiovascular risk factors,
alysed 2851 participants in the Dallas Heart Study and
including age, sex, WHR, systolic BP, diabetes, LDL choles-
showed that CKD was associated with CACS >100 in
terol, HDL cholesterol, hsCRP, urine PCR, statin use, smok-
the multivariate analysis. However, only 211 CKD pa-
ing status, serum calcium, serum phosphorus, 25‐OH‐vit D
tients were included, and it was impossible to analyse
HYUN et al.   
|
    5 of 8

T A B L E 2   Multivariate‐adjusted CACS ratios according to eGFR categories

eGFRMDRD eGFRCr eGFRCys eGFRCrCys

eGFR Ratios (95% CI) P Ratios (95% CI) P Ratios (95% CI) P Ratios (95% CI) P
Model 1
≥90 Reference   Reference   Reference   Reference  
60‐89 1.69 (0.75‐3.81) 0.206 2.08 (1‐4.32) 0.05 2.90 (1.37‐6.13) 0.005 2.36 (1.11‐5.02) 0.026
30‐59 2.82 (1.32‐6.01) 0.007 3.06 (1.56‐6) 0.001 3.44 (1.67‐7.06) 0.001 3.57 (1.77‐7.21) <0.001
<30 5.23 (2.37‐11.53) <0.001 5.53 (2.70‐11.33) <0.001 6.88 (3.31‐14.32) <0.001 6.09 (2.95‐12.6) <0.001
Model 2
≥90 Reference   Reference   Reference   Reference  
60‐89 1.60 (0.72‐3.55) 0.251 2.14 (1.05‐4.36) 0.036 2.62 (1.25‐5.47) 0.010 2.16 (1.03‐4.52) 0.041
30‐59 2.03 (0.95‐4.31) 0.066 2.03 (1.06‐3.89) 0.033 2.30 (1.13‐4.68) 0.022 2.36 (1.18‐4.69) 0.015
<30 2.71 (1.22‐6.05) 0.015 2.90 (1.41‐5.93) 0.004 3.44 (1.63‐7.25) 0.001 2.94 (1.40‐6.18) 0.004
Model 3
≥90 Reference   Reference   Reference   Reference  
60‐89 1.64 (0.73‐3.65) 0.227 2.18 (1.07‐4.46) 0.033 2.63 (1.26‐5.50) 0.010 2.15 (1.02‐4.50) 0.044
30‐59 2.03 (0.95‐4.34) 0.068 2.01 (1.04‐3.88) 0.038 2.23 (1.09‐4.56) 0.028 2.27 (1.13‐4.54) 0.021
<30 2.32 (1.01‐5.35) 0.048 2.45 (1.15‐5.23) 0.021 2.86 (1.30‐6.29) 0.009 2.39 (1.09‐5.22) 0.029
P for trend 0.040   0.033   0.018   0.034  
CAC, coronary artery calcification; CACS, coronary artery calcium score; CI, confidence interval; eGFR, estimated glomerular filtration rate; OR, odds ratio.
Model 1: adjusted for age and sex. Model 2: adjusted for model 1 + waist‐hip ratio, systolic blood pressure, diabetes, low‐density lipoprotein cholesterol, high‐density
lipoprotein cholesterol, high‐sensitivity C‐reactive protein, urine protein‐to‐creatinine ratio, statin use and current smoking. Model 3: adjusted for model 2 + calcium,
phosphorus, 25‐OH‐vit D and intact parathyroid hormone.

T A B L E 3   Multivariate‐adjusted ORs for CAC (CACS > 100) according to eGFR categories

eGFRMDRD eGFRCr eGFRCys eGFRCrCys

eGFR OR (95% CI) P OR (95% CI) P OR (95% CI) P OR (95% CI) P

Model 1
≥90 Reference   Reference   Reference   Reference  
60‐89 1.37 (0.66‐2.86) 0.400 1.41 (0.75‐2.68) 0.287 1.93 (0.97‐3.86) 0.062 2.04 (1.00‐4.18) 0.051
30‐59 2.08 (1.06‐4.11) 0.034 2.06 (1.15‐3.68) 0.015 2.57 (1.34‐4.92) 0.005 2.80 (1.43‐5.48) 0.003
<30 2.62 (1.32‐5.20) 0.006 2.64 (1.46‐4.76) 0.001 3.3 (1.72‐6.32) <0.001 3.58 (1.82‐7.01) <0.001
Model 2
≥90 Reference   Reference   Reference   Reference  
60‐89 1.43 (0.66‐3.07) 0.365 1.47 (0.76‐2.86) 0.255 2.15 (1.04‐4.43) 0.038 2.20 (1.04‐4.65) 0.038
30‐59 1.90 (0.93‐3.87) 0.078 1.77 (0.96‐3.25) 0.068 2.44 (1.23‐4.84) 0.011 2.63 (1.30‐5.32) 0.007
<30 2.13 (1.02‐4.46) 0.044 2.10 (1.11‐3.99) 0.023 2.86 (1.42‐5.79) 0.003 3.01 (1.46‐6.23) 0.003
Model 3 Reference   Reference   Reference   Reference  
≥90 Reference   Reference   Reference   Reference  
60‐89 1.40 (0.65‐3.02) 0.384 1.45 (0.74‐2.82) 0.278 2.10 (1.02‐4.32) 0.045 2.13 (1.01‐4.50) 0.047
30‐59 1.85 (0.91‐3.77) 0.091 1.72 (0.93‐3.18) 0.083 2.35 (1.19‐4.68) 0.014 2.51 (1.24‐5.10) 0.011
<30 1.88 (0.88‐4.03) 0.105 1.87 (0.95‐3.67) 0.069 2.52 (1.21‐5.23) 0.013 2.64 (1.24‐5.61) 0.012
P for 0.072   0.056   0.013   0.010  
trend
CAC, coronary artery calcification; CACS, coronary artery calcium score; CI, confidence interval; eGFR, estimated glomerular filtration rate; OR, odds ratio.
Model 1: adjusted for age and sex. Model 2: adjusted for model 1 + waist‐hip ratio, systolic blood pressure, diabetes, low‐density lipoprotein cholesterol, high‐density
lipoprotein cholesterol, high‐sensitivity C‐reactive protein, urine protein‐to‐creatinine ratio, statin use and current smoking. Model 3: adjusted for model 2 + calcium,
phosphorus, 25‐OH‐vit D and intact parathyroid hormone.
 |
6 of 8       HYUN et al.

T A B L E 4   Multivariate‐adjusted ORs (95% CI) for CAC with different cut‐offs

CACS > 10 (n = 589) CACS > 50 (n = 425) CACS > 200 (n = 240)

eGFRCrCys OR (95% CI) P OR (95% CI) P OR (95% CI) P

≥90 Reference   Reference   Reference  
60‐89 1.64 (0.99‐2.73) 0.056 1.49 (0.81‐2.74) 0.200 2.11 (0.86‐5.21) 0.104
30‐59 1.84 (1.14‐2.96) 0.013 2.02 (1.15‐3.56) 0.015 2.56 (1.09‐6.02) 0.031
<30 1.73 (1.01‐2.95) 0.046 1.99 (1.07‐3.69) 0.030 2.53 (1.02‐6.26) 0.045
P for trend 0.042   0.016   0.037  
CAC, coronary artery calcification; CACS, coronary artery calcium score; CI, confidence interval; eGFR, estimated glomerular filtration rate; OR, odds ratio.
Adjusted for age, sex, waist‐hip ratio, systolic blood pressure, diabetes, low‐density lipoprotein cholesterol, high‐density lipoprotein cholesterol, high‐sensitivity C‐reac-
tive protein, urine protein‐to‐creatinine ratio, statin use, current smoking, calcium, phosphorus, 25‐OH‐vit D and intact parathyroid hormone.

the relationship according to the level of renal function.
Two large studies have resulted from the CRIC (Chronic
Renal Insufficiency Cohort). In the analysis by Budoff
et al, 20 eGFR <30 mL/min/1.73 m 2 was associated with
CAC in multivariate analysis. However, this study did
not consider important risk factors, such as calcium or
phosphorus. In another analysis of the CRIC study by He
et al,21 high cystatin C level was associated with CACS
>100 in the multivariate analysis including calcium and
phosphorus. However, eGFR lost statistical significance
in the multivariate model.
Unlike the previous three studies using eGFRMDRD
for cross‐sectional analysis, Lamprea‐Montealegre et al22
performed a longitudinal analysis using eGFRCYS. They
analysed 2795 participants in the Multi‐Ethnic Study of
Atherosclerosis with no baseline CAC and showed that low
eGFR is associated with future CAC incidence. However, not
only did they miss important risk factors, such as obesity or
serum phosphorus, but also the association was attenuated
after adjustment for lipid phenotype. In this study, the gradual
association of eGFR and CAC was more prominent with the
formula using cystatin C. In particular, only the eGFRCrCys
improved the CAC prediction with statistical significance
in our study. This is in line with previous studies suggesting
that this formula has better predictions of real GFR and in-
dividual risk.15,23,24 This is mainly because serum creatinine
is affected by muscle mass. For this reason, KDIGO Clinical
Practice Guideline recommends eGFR using cystatin C be
used as a confirmatory test in specific circumstances such
as low muscle mass.25 One reason for the lack of relevance
of renal function and CAC in previous studies is presumably
the lower performance of the MDRD formula. As far as we
know, our study is the first to confirm a gradual and indepen-
dent association between eGFR and CAC even after adjust-
ment for comprehensive risk factors related to cardiovascular
disease.
The CAC observed in patients with CKD included both
intimal calcification and medial calcification.26 Therefore,
risk factors for intimal or medial calcification may be asso-
ciated with CAC in CKD. Intimal calcification is mainly as-
sociated with atherosclerosis and is expressed in the form of
a calcified atherosclerotic plaque.27 Traditional risk factors
for atherosclerosis, such as age, male gender, hypertension,
diabetes, obesity and smoking, are also risk factors for CKD
and are highly prevalent in CKD.28,29 In addition, pathophys-
iologic changes in CKD, such as abnormal levels of bone‐
related proteins, altered calcification‐regulating humoral
factors, chronic inflammation and endothelial dysfunction,
are inducers of medial calcification.27 In other words, the
risk factors for CKD overlap with those for CAC, and various
metabolic derangements in CKD cause CAC; hence, these
may explain the association between the low eGFR and CAC.
Our study has some limitations. First, we cannot confirm
a causal relationship between lower eGFR and CAC from this
cross‐sectional study. If prospective studies can confirm the
relationship between eGFR and the development or progres-
sion of CAC, it will be stronger evidence for the relationship
between eGFR and CAC. Second, GFR was not directly mea-
sured, but was estimated as eGFR. However, it is not possible
to use GFR in evaluating patients’ renal function in practice,
and thus, studies using eGFR may be more meaningful. In
addition, our use of data from a structured cohort study of
large numbers of CKD patients from multiple centres is a
strength of this study.
In conclusion, low eGFR was gradually associated with
higher CACS and higher prevalence of CAC in a predialy-
sis CKD cohort of Korean adults. This result was valid even
after adjustment for various traditional and nontraditional
cardiovascular risk factors, and in the analysis with a differ-
ent cut‐off for CAC. eGFR using both creatinine and cystatin
C significantly improved the prediction of CAC in CKD. Our
findings reaffirm the importance of preventing renal function
decline and the importance of controlling CV risk in CKD.
Future studies should focus on identifying more precise
mechanisms by which CKD induces CAC and CVD, as well
as proper interventions to improve CAC in CKD.
HYUN et al.
CONFLICT OF INTEREST
All the authors declared no conflict of interest.
AUTHOR CONTRIBUTION
Young Youl Hyun involved in drafting the article, analysis
and interpretation of data, revising the article and final ap-
proval of the article; Hyang Kim, Kook‐Hwan Oh, Curie
Ahn, Sue K. Park, Dong Wan Chae, Yun Kyu Oh, Kyu Hun
Choi, Seung Hyeok Han, Yeong Hoon Kim analysed and in-
terpreted the data, revised the article and involved in final ap-
proval of the article; Kyu‐Beck Lee involved in conception or
design, analysis and interpretation of data and final approval
of the article.
ORCID
Young Youl Hyun  https://orcid.
org/0000-0002-4204-9908
R E F E R E NC E S
1. Pletcher MJ, Tice JA, Pignone M, Browner WS. Using the cor-
onary artery calcium score to predict coronary heart disease
events: a systematic review and meta‐analysis. Arch Intern Med.
2004;164(12):1285‐1292.
2. Budoff MJ, Shaw LJ, Liu ST, et al. Long‐term prognosis associated
with coronary calcification: observations from a registry of 25,253
patients. J Am Coll Cardiol. 2007;49(18):1860‐1870.
3. Block GA, Raggi P, Bellasi A, Kooienga L, Spiegel DM. Mortality
effect of coronary calcification and phosphate binder choice in in-
cident hemodialysis patients. Kidney Int. 2007;71(5):438‐441.
4. Russo D, Corrao S, Battaglia Y, et al. Progression of coronary ar-
tery calcification and cardiac events in patients with chronic renal
disease not receiving dialysis. Kidney Int. 2011;80(1):112‐118.
5. Matsushita K, Sang Y, Ballew SH, et al. Subclinical atherosclerosis
measures for cardiovascular prediction in CKD. J Am Soc Nephrol.
2015;26(2):439‐447.
6. Chen J, Budoff MJ, Reilly MP, et al. Coronary artery calcification
and risk of cardiovascular disease and death among patients with
chronic kidney disease. JAMA Cardiol. 2017;2:635‐643.
7. Goodman WG, Goldin J, Kuizon BD, et al. Coronary‐artery calci-
fication in young adults with end‐stage renal disease who are un-
dergoing dialysis. N Engl J Med. 2000;342(20):1478‐1483.
8. Raggi P, Boulay A, Chasan‐Taber S, et al. Cardiac calcifica-
tion in adult hemodialysis patients. A link between end‐stage
renal disease and cardiovascular disease? J Am Coll Cardiol.
2002;39(4):695‐701.
9. Russo D, Palmiero G, De Blasio AP, Balletta MM, Andreucci VE.
Coronary artery calcification in patients with CRF not undergoing
dialysis. Am J Kidney Dis. 2004;44(6):1024‐1030.
10. Kramer H, Toto R, Peshock R, Cooper R, Victor R. Association
between chronic kidney disease and coronary artery calcification:
the Dallas Heart Study. J Am Soc Nephrol. 2005;16(2):507‐513.
11. Levey AS, Coresh J, Greene T, et al. Using standardized serum
creatinine values in the modification of diet in renal disease study
  
|
    7 of 8
equation for estimating glomerular filtration rate. Ann Intern Med.
2006;145(4):247‐254.
12. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate
glomerular filtration rate. Ann Internal Med. 2009;150(9):604‐612.
13. Matsushita K, Mahmoodi BK, Woodward M, et al. Comparison
of risk prediction using the CKD‐EPI equation and the MDRD
study equation for estimated glomerular filtration rate. JAMA.
2012;307(18):1941‐1951.
14. Oh K‐H, Park SK, Park HC, et al. KNOW‐CKD (KoreaN cohort
study for Outcome in patients With Chronic Kidney Disease): de-
sign and methods. BMC Nephrol. 2014;15:80.
15. Inker LA, Schmid CH, Tighiouart H, et al. Estimating glomerular
filtration rate from serum creatinine and cystatin C. N Engl J Med.
2012;367(1):20‐29.
16. Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte
M Jr, Detrano R. Quantification of coronary artery calcium
using ultrafast computed tomography. J Am Coll Cardiol.
1990;15(4):827‐832.
17. Raggi P, Callister TQ, Cooil B, et al. Identification of patients
at increased risk of first unheralded acute myocardial infarc-
tion by electron‐beam computed tomography. Circulation.
2000;101(8):850‐855.
18. Berman DS, Wong ND, Gransar H, et al. Relationship between
stress‐induced myocardial ischemia and atherosclerosis mea-
sured by coronary calcium tomography. J Am Coll Cardiol.
2004;44(4):923‐930.
19. Simera I, Moher D, Hoey J, Schulz KF, Altman DG. A catalogue
of reporting guidelines for health research. Eur J Clin Invest.
2010;40(1):35‐53.
20. Budoff MJ, Rader DJ, Reilly MP, et al. Relationship of es-
timated GFR and coronary artery calcification in the CRIC
(Chronic Renal Insufficiency Cohort) Study. Am J Kidney Dis.
2011;58(4):519‐526.
21. He J, Reilly M, Yang W, et al. Risk factors for coronary artery
calcium among patients with chronic kidney disease (from the
Chronic Renal Insufficiency Cohort Study). Am J Cardiol.
2012;110(12):1735‐1741.
22. Lamprea‐Montealegre JA, McClelland RL, Astor BC, et al. Chronic
kidney disease, plasma lipoproteins, and coronary artery calcium
incidence: the multi‐ethnic study of atherosclerosis. Arterioscler
Thromb Vasc Biol. 2013;33(3):652‐658.
23. Meeusen JW, Rule AD, Voskoboev N, Baumann NA, Lieske JC.
Performance of cystatin C‐ and creatinine‐based estimated glo-
merular filtration rate equations depends on patient characteristics.
Clin Chem. 2015;61(10):1265‐1272.
24. Rogacev Ks, Pickering Jw, Seiler S, et al. The Chronic Kidney
Disease Epidemiology Collaboration (CKD‐EPI) equation incor-
porating both cystatin C and creatinine best predicts individual
risk: a cohort study in 444 patients with chronic kidney disease.
Nephrol Dial Transplant. 2014;29(2):348‐355.
25. KDIGO. KDIGO clinical practice guideline for the evaluation and
management of chronic kidney disease; 2012.
26. Nakamura S, Ishibashi‐Ueda H, Niizuma S, Yoshihara F, Horio
T, Kawano Y. Coronary calcification in patients with chronic kid-
ney disease and coronary artery disease. Clin J Am Soc Nephrol.
2009;4(12):1892‐1900.
27. Vervloet M, Cozzolino M. Vascular calcification in chronic
kidney disease: different bricks in the wall? Kidney Int.
2017;91(4):808‐817.
 |
8 of 8       HYUN et al.

28. Fruchart J‐C, Nierman MC, Stroes E, Kastelein J, Duriez P.

New risk factors for atherosclerosis and patient risk assessment. How to cite this article: Hyun YY, Kim H, Oh K‐H,
Circulation. 2004;109(23 suppl 1):III‐15‐III‐19. et al. eGFR and coronary artery calcification in
29. Kazancioğlu R. Risk factors for chronic kidney disease: an update. chronic kidney disease. Eur J Clin Invest.
Kidney Int Suppl. 2013;3(4):368‐371. 2019;49:e13101. https://doi.org/10.1111/eci.13101

SUPPORTING INFORMATION
Additional supporting information may be found online in
the Supporting Information section at the end of the article.